New try last version

Not a GPU necessary anymore
Different cover styles possible

app.py

@@ -1,33 +1,79 @@
+# -*- coding: utf-8 -*-
+"""20221027_Generate_Beatles_with_Gradio_faster_version.ipynb
+
+Automatically generated by Colaboratory.
+
+Original file is located at
+    https://colab.research.google.com/drive/1spCrOvhSvByMN8J-4iGZ9wk3I8U-9elU
+
+# Build a demo
+
+Put together a Gradio App bases on the Huggingface tutorials so that users can generate Beatles-like poetry based on an input prompt.
+"""
+
+# Commented out IPython magic to ensure Python compatibility.
+# %%capture
+# !pip install transformers
+
+# Commented out IPython magic to ensure Python compatibility.
+# %%capture
+# !pip install gradio
+
+# Commented out IPython magic to ensure Python compatibility.
+# %%capture
+# !pip install --upgrade diffusers transformers scipy ftfy "ipywidgets>=7,<8"
+
+# Commented out IPython magic to ensure Python compatibility.
+# %%capture
+# !pip install flag
+
+# Import libraries
 from transformers import pipeline
+from numpy import random
 import gradio as gr
 import re
-import os
 import torch
 from torch import autocast
-from diffusers import StableDiffusionPipeline
+#from diffusers import StableDiffusionPipeline
+from PIL import Image
+import flag
+import os
+
+image_input_styles = ["Random", "Beatles style", "Pencil sketch", "Oil painting", "Pop art", "Piet Mondriaan"]
+
+# Function to get random image type for image input
+def get_image_input(title, given_input_style):
+    if given_input_style == 'Random':
+      image_input_styles_new = image_input_styles.copy()
+      image_input_styles_new.pop(0)
+      random_choice = random.randint(len(image_input_styles_new)-1)
+      final_style = image_input_styles_new[random_choice]
+    else:
+      final_style = given_input_style
+    image_input = final_style + ' for ' + title
+    return image_input, final_style
 
-# Available models for pipeline
-# checkpoint = 'wvangils/CTRL-Beatles-Lyrics-finetuned-newlyrics'
-# checkpoint = 'wvangils/GPT-Medium-Beatles-Lyrics-finetuned-newlyrics'
-checkpoint = 'wvangils/GPT-Neo-125m-Beatles-Lyrics-finetuned-newlyrics'
-# checkpoint = 'wvangils/GPT2-Beatles-Lyrics-finetuned-newlyrics'
-# checkpoint = 'wvangils/DistilGPT2-Beatles-Lyrics-finetuned-newlyrics'
-# checkpoint = 'wvangils/BLOOM-350m-Beatles-Lyrics-finetuned-newlyrics'
+# Available models for generate lyrics pipeline
+checkpoint = 'wvangils/GPT-Medium-Beatles-Lyrics-finetuned-newlyrics'
+# checkpoint = 'wvangils/GPT-Neo-125m-Beatles-Lyrics-finetuned-newlyrics'
+# checkpoint = 'wvangils/BLOOM-560m-Beatles-Lyrics-finetuned'
 
-# setup all the pipelines we need
+# Setup all the pipelines we need
 title_generator = pipeline('summarization', model='czearing/story-to-title')
 lyrics_generator = pipeline("text-generation", model=checkpoint)
-image_generator = StableDiffusionPipeline.from_pretrained("CompVis/stable-diffusion-v1-4"
-                                                , revision="fp16"
-                                                , torch_dtype=torch.float16
-                                                , use_auth_token=os.environ['hf_token'])  
+# Image generator: stable diffusion from an existing HuggingFace space
+stable_diffusion = gr.Blocks.load(name="spaces/stabilityai/stable-diffusion")
 
-# What device to use?
-device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
-image_generator.to(device)
+# Create 4 images for the given prompt and receive the first one
+ # This function uses an existing HuggingFace space where the number of created images cannot be modified
+def get_image(prompt):
+    gallery_dir = stable_diffusion(prompt, fn_index=2)
+    images = [os.path.join(gallery_dir, img) for img in os.listdir(gallery_dir)]
+    return [images[0]]
 
-# Create function for generation
-def generate_beatles(input_prompt, temperature, top_p):
+# Create function for lyrics generation
+def generate_beatles(input_prompt, temperature, top_p, given_input_style):
+    # Create generator for different models
     generated_lyrics = lyrics_generator(input_prompt
                               , max_length = 100
                               , num_return_sequences = 1
@@ -42,55 +88,55 @@ def generate_beatles(input_prompt, temperature, top_p):
                               #, do_sample = True     # Default = False
                               )[0]["generated_text"]
 
+    # Put lyrics in the right form
     lyrics_sentences = re.sub('\n', '. ', generated_lyrics)
-    #title_generator = pipeline('summarization', model='czearing/story-to-title')
+
+    # Create a title based on the generated lyrics
     title = title_generator(lyrics_sentences, min_length=1, max_length=10, repetition_penalty=2.5)[0]['summary_text']
 
-    # Set a seed for reproducebility
-    # seed = torch.Generator("cuda").manual_seed(42)
-    image_input = "Oil painting for " + title
+    # Create an image based on the generated title
+    image_input, image_style = get_image_input(title, given_input_style)
 
     # Generate the image, [PIL format](https://pillow.readthedocs.io/en/stable/)
-    with autocast("cuda"):
-      image = image_generator(image_input, num_inference_steps=50, guidance_scale=8.5)[0][0] 
-
-    return (title, generated_lyrics, image)
-
-
-# Create textboxes for input and output
-input_box = gr.Textbox(label="Input prompt:", placeholder="Write the start of a song here", value="In my dreams I am", lines=2, max_lines=5)
-gen_lyrics = gr.Textbox(label="Lyrics by The Beatles and chosen language model:", lines=15)
-gen_title = gr.Textbox(label="Proposed songtitle", lines=1) 
-gen_image = gr.Image(label="Proposed song cover", type="pil")
+    image = get_image(image_input)
+    return (title, generated_lyrics, image, image_style)
 
 # Create textboxes for input and output
-input_box = gr.Textbox(label="Input prompt:", placeholder="Write the start of a song here", value="In my dreams I am", lines=2, max_lines=5)
-gen_lyrics = gr.Textbox(label="Lyrics by The Beatles and chosen language model:", lines=15)
-gen_title = gr.Textbox(label="Proposed songtitle", lines=1) 
-gen_image = gr.Image(label="Proposed song cover", type="pil")
+input_box = gr.Textbox(label="Write the start of a song here", placeholder="Write the start of a song here", value="Looking beyond the Broad Horizon", lines=2, max_lines=5)
+gen_lyrics = gr.Textbox(label="Lyrics by The Beatles and chosen language model", lines=15)
+gen_title = gr.Textbox(label="Proposed songtitle", lines=1)
+gen_image = gr.Gallery(label="Proposed song cover").style(grid=1, height="auto")
+gen_image_style = gr.Textbox(label="Image style", lines=1)
 
 # Layout and text above the App
 title='Beatles lyrics generator'
-description="<p style='text-align: center'>We've fine-tuned multiple language models on lyrics from The Beatles to generate Beatles-like text. Below are the results we obtained fine-tuning a GPT neo model. After generation a title is generated using <a href='https://huggingface.co/czearing/story-to-title' target='_blank'>this model</a>. On top we use the generated title to suggest an album cover using <a href='https://huggingface.co/CompVis/stable-diffusion-v1-4' target='_blank'>Stable Diffusion 1.4</a>. Give it a try!</p>"
+description="<p style='text-align: center'>We've fine-tuned multiple language models on lyrics from The Beatles to generate Beatles-like text. Below are the results we obtained fine-tuning a GPT Neo model. After generation a title is generated using <a href='https://huggingface.co/czearing/story-to-title' target='_blank'>this model</a>. On top we use the generated title to suggest an album cover using <a href='https://huggingface.co/CompVis/stable-diffusion-v1-4' target='_blank'>Stable Diffusion 1.4</a>. Give it a try!</p>"
 article="""<p style='text-align: left'>These text generation models that output Beatles-like text were created by data scientists working for <a href='https://cmotions.nl/' target="_blank">Cmotions.</a>  
         We tried several text generation models that we were able to load in Colab: a general <a href='https://huggingface.co/gpt2-medium' target='_blank'>GPT2-medium</a> model, the Eleuther AI small-sized GPT model <a href='https://huggingface.co/EleutherAI/gpt-neo-125M' target='_blank'>GPT-Neo</a> and the new kid on the block build by the <a href='https://bigscience.notion.site/BLOOM-BigScience-176B-Model-ad073ca07cdf479398d5f95d88e218c4' target='_blank'>Bigscience</a> initiative <a href='https://huggingface.co/bigscience/bloom-560m' target='_blank'>BLOOM 560m</a>. 
         Further we've put together a <a href='https://huggingface.co/datasets/cmotions/Beatles_lyrics' target='_blank'> Huggingface dataset</a> containing all known lyrics created by The Beatles. Currently we are fine-tuning models and are evaluating the results. Once finished we will publish a blog at this <a href='https://www.theanalyticslab.nl/blogs/' target='_blank'>location </a> with all the steps we took including a Python notebook using Huggingface.
         The default output contains 100 tokens and has a repetition penalty of 1.0. 
          </p>"""
-         
+css = """
+        button:lang(en-US) {
+        }  
+""" 
 
 # Let users select their own temperature and top-p
-temperature = gr.Slider(minimum=0.1, maximum=1.0, step=0.1, label="Temperature (high = sensitive for low probability tokens)", value=0.7, show_label=True)
-top_p = gr.Slider(minimum=0.1, maximum=1.0, step=0.1, label="Top-p (sample next possible words from given probability p)", value=0.5, show_label=True)
+temperature = gr.Slider(minimum=0.1, maximum=1.0, step=0.1, label="Change the temperature \n (higher temperature = more creative in lyrics generation, but posibbly less Beatly)", value=0.5, show_label=True)  #high = sensitive for low probability tokens
+top_p = gr.Slider(minimum=0.1, maximum=1.0, step=0.1, label="Change top probability of the next word \n (higher top probability = more words to choose from for the next word, but possibly less Beatly)", value=0.5, show_label=True)
+given_input_style = gr.Dropdown(choices=image_input_styles, value="Random", label="Choose the art style for the lyrics cover", show_label=True)
 #checkpoint = gr.Radio(checkpoint_choices, value='wvangils/GPT-Medium-Beatles-Lyrics-finetuned-newlyrics', interactive=True, label = 'Select fine-tuned model', show_label=True)
 
 # Use generate Beatles function in demo-app Gradio
 gr.Interface(fn=generate_beatles
-             , inputs=[input_box, temperature, top_p]
-             , outputs=[gen_title, gen_lyrics, gen_image]
+             , inputs=[input_box, temperature, top_p, given_input_style]
+             , outputs=[gen_title, gen_lyrics, gen_image, gen_image_style]
              , title=title
+             , css=css
              , description=description
              , article=article
              , allow_flagging='never'
-             ).launch()
-             
+             ).launch(debug=True, share=True)
+
+#generate_beatles(input_prompt="When I look out my window", temperature=0.7, top_p=0.5)
+